Shinyi Wu - US grants

1548502(Mataric)

The number of adults over the age of 65 is predicted to reach 83.7 million by 2050. There is accordingly a growing need for technologies for the elderly to complement human care in addressing the fast-growing needs of the elderly population. This project takes a novel approach to developing and evaluating an assistive technology for the elderly: it views the family unit as a team, and focuses on developing a robots as a team member (rather than individual assistant) designed to aid in the achievement of team/family goals. In the project, these goals are focused on the wellbeing of the elderly family member, and include daily social connectedness with others in the family, regular physical and cognitive activity, and medication adherence. These goals are part of the family dynamic and involve the interaction of various team members, pairwise, in subgroups, or all together. The project will develop a hardware and software infrastructure for deploying a socially assistive robot in the home and methods for enabling the robot to adapt to the family members in order to aid in facilitating team dynamics and aiding family goal achievement. The project brings together leading experts in socially assistive robotics, gerontology, and social work, who will spend two years in a tightly knit process of technology development combined with in home system evaluation with participating families.

This project is focused on engineering team interactions and dynamics, with the robot as a member of the team, embedded in a real-world environment and interacting with other team members in real time. The engineering goal is to develop data-driven robot control algorithms and the associated hardware and software infrastructure for enabling robots to serve as team members in the context of family-supported eldercare. The control algorithms, combined with the hardware and software infrastructure, will enable the robot system to perceive and understand the relevant aspects of the family interaction dynamics, be a participant/component in those team dynamics, and influence/streer those team dynamics toward desired family team goals. The project will enable and then iteratively probe the effects of the robot as team member in different interaction contexts (user-family, user-caregiver, family-caregiver). Furthermore, the project tackles three fundamental components in team goal achievement: evaluation of team performance, coordination of actions across multiple team members, and team member role assignment. It also investigates the application of unsupervised, supervised, and reinforcement learning techniques to the domain of elderly care. The research is complemented with a K-12 STEM outreach program consisting of annual free interactive workshops for students in secondary education (grades 6-12; up to 30 students per workshop). Simple robot programming and demonstrations of the hardware and software being developed in the project will be used to discuss how assistive robots could play a role in society, starting with the students' families.